Console Output

Training and evaluating model for: Tablet
Dataset length: 62229 windows


 NILMModel(
  (conv1d): Conv1d(9, 9, kernel_size=(3,), stride=(1,), padding=(1,))
  (lstm): LSTM(9, 256, num_layers=4, batch_first=True, dropout=0.1)
  (dropout): Dropout(p=0.1, inplace=False)
  (relu): ReLU()
  (output_layer): Linear(in_features=256, out_features=1, bias=True)
)
Epoch [1/300], Train Loss: 0.017794
Validation Loss: 0.009140
Epoch [2/300], Train Loss: 0.008862
Validation Loss: 0.008334
Epoch [3/300], Train Loss: 0.008386
Validation Loss: 0.008229
Epoch [4/300], Train Loss: 0.008153
Validation Loss: 0.007944
Epoch [5/300], Train Loss: 0.008014
Validation Loss: 0.008214
Epoch [6/300], Train Loss: 0.007819
Validation Loss: 0.007780
Epoch [7/300], Train Loss: 0.007659
Validation Loss: 0.007466
Epoch [8/300], Train Loss: 0.007504
Validation Loss: 0.007303
Epoch [9/300], Train Loss: 0.007223
Validation Loss: 0.006868
Epoch [10/300], Train Loss: 0.007054
Validation Loss: 0.006909
Epoch [11/300], Train Loss: 0.006770
Validation Loss: 0.007016
Epoch [12/300], Train Loss: 0.006636
Validation Loss: 0.006502
Epoch [13/300], Train Loss: 0.006596
Validation Loss: 0.006527
Epoch [14/300], Train Loss: 0.006685
Validation Loss: 0.006845
Epoch [15/300], Train Loss: 0.006583
Validation Loss: 0.006390
Epoch [16/300], Train Loss: 0.006774
Validation Loss: 0.006403
Epoch [17/300], Train Loss: 0.006549
Validation Loss: 0.006293
Epoch [18/300], Train Loss: 0.006671
Validation Loss: 0.006363
Epoch [19/300], Train Loss: 0.009658
Validation Loss: 0.015352
Epoch [20/300], Train Loss: 0.015383
Validation Loss: 0.014995
Epoch [21/300], Train Loss: 0.013984
Validation Loss: 0.007340
Epoch [22/300], Train Loss: 0.007495
Validation Loss: 0.007009
Epoch [23/300], Train Loss: 0.007204
Validation Loss: 0.006865
Epoch [24/300], Train Loss: 0.007094
Validation Loss: 0.007077
Epoch [25/300], Train Loss: 0.007404
Validation Loss: 0.006866
Epoch [26/300], Train Loss: 0.006716
Validation Loss: 0.006779
Epoch [27/300], Train Loss: 0.006606
Validation Loss: 0.006244
Epoch [28/300], Train Loss: 0.006508
Validation Loss: 0.006225
Epoch [29/300], Train Loss: 0.006355
Validation Loss: 0.006424
Epoch [30/300], Train Loss: 0.006250
Validation Loss: 0.006412
Epoch [31/300], Train Loss: 0.006275
Validation Loss: 0.006127
Epoch [32/300], Train Loss: 0.006271
Validation Loss: 0.006137
Epoch [33/300], Train Loss: 0.006498
Validation Loss: 0.013372
Epoch [34/300], Train Loss: 0.009427
Validation Loss: 0.006853
Epoch [35/300], Train Loss: 0.006711
Validation Loss: 0.006729
Epoch [36/300], Train Loss: 0.006450
Validation Loss: 0.006353
Epoch [37/300], Train Loss: 0.007043
Validation Loss: 0.006852
Epoch [38/300], Train Loss: 0.006578
Validation Loss: 0.006283
Epoch [39/300], Train Loss: 0.006348
Validation Loss: 0.006179
Epoch [40/300], Train Loss: 0.006519
Validation Loss: 0.006451
Epoch [41/300], Train Loss: 0.006292
Validation Loss: 0.006212
Early stopping triggered

Evaluating model for: Tablet
Validation MAE: 0.613805 W
Validation MSE: 0.963927 W²
Validation RMSE: 0.981798 W
Signal Aggregate Error (SAE): 0.002398
Normalized Disaggregation Error (NDE): 0.225844

Training and Validation Loss

Training Loss Plot

Interactive Plot